SV40 DNA Contamination in COVID-19 mRNA Vaccines: The Definitive Investigation

Investigation Protocol: Every claim is tied to primary sources including regulator emails, peer-reviewed studies, manufacturer patents, and laboratory data. Quantitative values preserve original units and methodological caveats.

Executive summary#

• 8 independent laboratories across 4 continents have confirmed plasmid DNA contamination in mRNA COVID-19 vaccines, including lots with expiry dates extending into 2026
• The regulators' own record corroborates presence: TGA laboratory testing of 28 batches, PEI vial testing, Health Canada's written SV40 admission, and the EMA's BWP position paper. The open dispute is quantity and significance
• SV40 promoter/enhancer sequences present in all positive samples, with some lots exceeding regulatory DNA limits by 500-fold
• Health Canada internal emails reveal regulatory crisis: Pfizer admitted unexpected SV40 finding, officials noted "SV40 must be avoided!" and sought to "remedy the situation before Fall 2024"
• The EMA's own released position paper (EMA/CHMP/BWP/61303/2023) confirms in writing that all four SV40 elements are present in the BNT162b2 plasmid starting material, that the applicant did not highlight them at original authorisation, and that the compliance figure is measured on Drug Substance, not the injected Drug Product
• Achs et al. (npj Vaccines, December 2025), cited to regulators as reassurance, is contradicted by its own sequencing deposit: near-complete E. coli genomic DNA in four of five Comirnaty lots (1.4-2.4 percent of all reads), whole circular plasmid in every lot, and a fragment-length signature consistent with RNA:DNA hybrid protection
• The SV40 enhancer functions as a somatic hypermutation targeting element, peer-reviewed in 2024. The target is the antibody-diversity machinery. What that would mean in vaccinated people is unstudied.
• Manufacturer prior knowledge: Moderna's patent US 10,077,439 B2 (priority date 2013) states residual DNA "has the potential to be oncogenic in patient populations" and that qPCR does not measure "all other smaller DNA molecules that are partially digested"
• Regulatory transparency failure: No public comparability data for Process 1 vs Process 2; safety comparison cohort removed from trials

Eight laboratories, the same finding#

Between February 2023 and December 2024, eight independent research groups across North America, Europe, and Asia reported identical findings: mRNA COVID-19 vaccine vials contain residual plasmid DNA with functional SV40 promoter/enhancer sequences.

Different methods, different continents, same result. That convergence is what makes the finding hard to dismiss.

Laboratory Findings at a Glance#

All measurements normalized to 0.3 mL dose where applicable; methodologies vary

The regulators' own record corroborates presence (added 4 October 2026)#

The confirmation does not rest on the independent laboratories alone. Australia's TGA tested 28 batches in its own laboratories and published a summary report on residual DNA and endotoxin; its public position disputes the word "excess", not the presence of residual DNA. Germany's Paul Ehrlich Institute tested vials in December 2023 and reported no replication-competent plasmid DNA. Health Canada confirmed in writing that the SV40 promoter/enhancer is present in the vials (see the Health Canada section). The EMA's own BWP position paper, released under ASK-289849, confirms all four SV40 elements are present in the plasmid starting material (see the EMA section).

Between the laboratories and the regulators, the presence of residual plasmid DNA is no longer in dispute. What is disputed is quantity, fragment size, and biological significance. That is where the independent measurements and the regulators' compliance figures diverge, and that divergence is the subject of the qPCR, fluorometry, and Achs sections above.

February 2023: the first finding#

Kevin McKernan, a genomics researcher with decades of experience in DNA sequencing, was testing mRNA COVID-19 vaccines in February 2023 when he found plasmid DNA fragments carrying SV40 sequences in the finished vials. That genetic material should not be in a finished mRNA product. In some lots the levels ran hundreds of times above the established safety limit.

The limit itself is old. The International Council for Harmonisation caps residual DNA at 10 nanograms per dose, a guideline regulators worldwide have applied for decades. Multiple independent laboratories would later find levels far above it.

The RNA:DNA Hybrid Mechanism: Why DNA Persists Despite "Purification"#

The fundamental reason DNA contamination persists in mRNA vaccines is a molecular biology phenomenon that manufacturers failed to address: RNA:DNA hybrids resist standard DNase I digestion.

The Hybrid Formation Process#

During in vitro transcription (IVT), the process that copies mRNA from its DNA template, the newly made mRNA does not fully separate from the template. The two stay bound as an RNA:DNA hybrid on the plasmid.

Why Standard Clean-Up Failed#

The DNase I Problem:

• Standard purification uses DNase I enzyme to degrade residual DNA
• DNase I does not cleave the DNA strand of RNA:DNA hybrid duplexes (Sutton et al., 1997, PMID: 9020884)
• N1-methylpseudouridine modifications in therapeutic mRNA further stabilize these hybrids

The selective protection effect:

• Hybrid-protected regions: the spike gene ORF, where mRNA binds DNA, resists DNase I
• Non-hybridized regions: the kanamycin resistance gene and the origin of replication are degraded by DNase I

The Regulatory Assay Blind Spot#

This is how the assay misses it:

1. Manufacturers test only the kanamycin resistance gene with qPCR (assuming it represents total DNA)
2. Kanamycin gene shows low levels because it's degraded by DNase I
3. Regulators see "acceptable" levels and approve the batch
4. Hidden contamination persists in the spike gene region protected by hybrids

Key Paper: McKernan et al. (2025) demonstrated this mechanism across 5 recent vaccine lots, showing ΔCt values of 6-15 between spike and kanamycin assays - representing 100-1000-fold more DNA than regulators detected.

Figure 1: The 100-fold discrepancy. 27 qPCRs per vial (3 targets × 3 dilutions × triplicates). Fluorometry sees the real DNA load. KAN qPCR sees almost nothing. Switch to Spike/SV40/Ori targets and the numbers match.

Source: Kevin McKernan, DNA contamination analysis, 2024 - Fluorometry using 1000× excess of RNaseA shows qPCR vs total DNA discrepancy

The same body of work quantified it directly: RNase A fluorometry of two Pfizer lots and three Moderna lots put residual DNA 15-48 times above the 10 ng per dose limit in every preparation tested.

Figure 2: Five lots, every preparation over the line. RNase A-Qubit fluorometry of two Pfizer lots (LN2588, GK0936) and three Moderna lots (025G23A, AW4694B, AT0709B), tested neat, after Triton X-100, after 95 C heat, and after RNase A. Every preparation reads above the 10 ng/dose FDA/WHO limit, most by one to three orders of magnitude.

Source: McKernan K, Rixey C, Rose J, "RNA:DNA hybrids survive digestion in mRNA vaccine manufacturing", Zenodo 2025, doi:10.5281/zenodo.17832183, Figure 2.

DNase I-XT (an enhanced enzyme cocktail) can effectively degrade RNA:DNA hybrids:

• Reduces DNA levels by 100-1000x compared to standard DNase I
• Shifts spike qPCR Ct values by 10-15 cycles
• Recommended for mRNA manufacturing but not implemented by Pfizer/Moderna

Figure 3: The fix that was not used. Standard DNase leaves Ct ~20-25, which means a lot of protected DNA. DNaseI-XT shifts the signal 6-12 cycles later; it actually cleans the preparation. The tool existed and was not deployed on the right target.

Source: McKernan et al., Zenodo (2025), doi:10.5281/zenodo.17832183 - DNaseI vs DNaseI-XT rescue curves showing Ct shift

~5.3 kb Nanopore Reads: The Physical Proof#

Figure 4: A single ~5.3 kb read. The fragments come straight from the vials; the highlighted blue section is one Nanopore read covering the spike region. That is intact, functional-length DNA, not "trace".

Source: McKernan et al., Nanopore sequencing analysis, ~5.3 Kb Nanopore Read showing physical proof of intact Spike fragments

A read this long shows the contamination is not limited to tiny fragments. DNA of this length is capable of:

• Complete SV40 promoter + Spike ORF expression
• Functional integration via NHEJ or LINE-1
• Episomal persistence in nucleus

The complete plasmid map#

Figure 5: The proposed inhibition mechanisms. McKernan's map overlays G-quadruplex structures and GAA repeats on the spike coding region. His proposal: these secondary structures further protect the spike DNA from degradation and help explain why spike signal runs 100× higher than KAN in qPCR.

Source: McKernan et al., follow-up analysis showing G-quadruplex formation in Spike region contributes to protection

[Calibration note, 4 October 2026:] a scan of OR134577.1 run by this author found zero canonical G-quadruplex motifs anywhere in the plasmid; the G-tracts on the minus strand are individual runs, not folded G4 structures. The G4-protection reading above should be treated as McKernan's proposal, not verified sequence fact. The verified explanation for the spike-versus-KAN asymmetry is the RNA:DNA hybrid protection documented above, now independently measured in the Achs deposit fragment lengths.

McKernan's annotated map of the BNT162b2 production plasmid, the public sequence now deposited as GenBank OR134577.1. The full eleven-element annotation from this author's independent re-annotation is in The BNT162b2 Plasmid Map.

Implications for All Reported Findings#

This hybrid mechanism explains:

• Why fluorometry shows 100x more DNA than qPCR (fluorometry detects all DNA after RNase treatment)
• Why spike gene persists while kanamycin gene is degraded
• Why SV40 promoters survive (located in protected regions)
• Why regulatory tests passed while contamination remained severe

Additional evidence: Kevin McKernan's sequencing work shows RNA/DNA hybrids resist enzymatic digestion, so the fragments may stay intact longer than standard degradation assays predict.

Manufacturer confirmed: BioNTech's 2024 paper#

This mechanism is documented in the manufacturer's own research, not only in the critics' data.

Lenk et al. (BioNTech, 2024) Frontiers in Molecular Biosciences:

"The specific activity of DNase I for RNA:DNA hybrids, however, is at least 100-fold below that for dsDNA (Sutton et al., 1997)."

Figure: the manufacturer's own number. "The specific activity of DNase I for RNA:DNA hybrids, however, is at least 100-fold below that for dsDNA." (Lenk et al., BioNTech 2024.) The enzyme the process relies on is, by BioNTech's own citation, at least 100-fold weaker on exactly the hybrids the process creates. Regulators still only checked the KAN gene that does get digested.

Source: Frontiers in Molecular Biosciences, DOI: 10.3389/fmolb.2024.1426129

The Lenk paper states that:

• RNA:DNA hybrids form during IVT
• Standard DNase I cannot effectively degrade them
• The protection effect is at least 100-fold
• Alternative purification methods are required

The manufacturer's own scientists documented the mechanism that enables DNA contamination to persist. Standard DNase I purification stayed in place, and the regulatory qPCR assays targeted the KAN gene, which gets destroyed, while the protected spike-encoding DNA survives. The persistence argument now rests on the manufacturer's own literature, not only on the independent laboratories.

Source: Lenk et al., "Process and analytical strategies for the safe production of mRNA vaccines and therapeutics," Frontiers in Molecular Biosciences, 2024. DOI: 10.3389/fmolb.2024.1426129

Full paper: https://www.frontiersin.org/journals/molecular-biosciences/articles/10.3389/fmolb.2024.1426129/full

The Regulator-Cited Reassurance Paper, Re-Mapped: Achs et al. (added 4 October 2026)#

On 13 December 2025, npj Vaccines published Achs et al., "Systematic analysis of COVID-19 mRNA vaccines using four orthogonal approaches demonstrates no excessive DNA impurities". The title is the conclusion. It has since been cited to regulators as reassurance that residual DNA in Comirnaty and Spikevax is within limits, the fragments are small, and nothing derives from host genomes.

Two things have happened since publication.

The method critique, now peer-reviewed#

McKernan, Speicher and Rose published "Systematic methodological flaws in DNA contamination assessment of mRNA vaccines: a critical analysis of Achs et al. (2025)" in Oncotarget (2026;17:368-376). The critique is theirs; this summary adds nothing to it. The eight points in brief:

1. qPCR amplicons sized 63 bp (KAN) to 233 bp (SPIKE) against a fragment population the paper's own data put at a 130-201 bp median. The 233 bp spike target is mostly unamplifiable, and the spike:backbone 4-6 Ct offset seen in four independent laboratories (McKernan, Speicher, Buckhaults, Fleming) disappears into a size artifact.
2. Copy-to-mass conversion applied full-plasmid molecular weight to a fragmented single-target count. The only coherent readings are roughly 0.07 ng/dose (fragment mass) or 12.2 ng/dose (fragmentation-corrected, over the 10 ng limit). The published 8 ng sits between the two.
3. No fragmentation correction anywhere in the qPCR chain, even where the paper's own median fragment length requires one.
4. Fluorometry switched to a DNA-purification front end no prior laboratory used, with the methods edited post hoc (ladder changed, no recovery number moved).
5. Capillary electrophoresis with electrokinetic injection, no matrix controls, standards in clean buffer. An LNP matrix that cannot inject reads as "below detection" regardless of true content.
6. The CE instrument's smear floor is 50 pg/uL. The 10 ng per 300 uL dose limit equals 33.3 pg/uL. The instrument is blind across the entire band where compliance is decided.
7. Fragment sizing done on Illumina after 95 C for 10 min, two SPRI cleanups and 10 PCR cycles, in a laboratory that owns an ONT instrument. The measured medians describe the library prep, not the vials.
8. Qubit trusted to titrate DNA input for the paper's own Illumina runs, and dismissed as unreliable when other laboratories reported over-limit DNA with it. Both statements appear in the same paper.

Jessica Rose's Lander-Waterman treatment of the same data reaches the same place by a different road: detected KAN copies represent roughly 37 percent of the true KAN-bearing fragment population at conservative boundary assumptions, which puts the corrected mass around 25 ng per dose. Every coherent correction lands above the 10 ng limit.

What the paper's own deposit shows#

The sequencing data behind the paper are public: NCBI BioProject PRJNA1303864. This author re-mapped the fifteen deposited libraries with open tools (minimap2, samtools), predictions registered in the script headers before each run. The full analysis, the run table and the standing caveats are in the companion article, The Host Genome They Said Was Not There. Four findings bear directly on this article.

1. The host genome the paper said was not there. In four of five Comirnaty lots, 1.4-2.4 percent of ALL sequencing reads map across 96 percent of the complete E. coli K12 genome at 56-95x mean depth. That is 13,700 to 23,700 reads per million. All six Spikevax lots sit at 8-472 rpm. Every Comirnaty lot is above every Spikevax lot, a 3,100x brand split. A shared-reagent kitome cannot produce a brand split, and the phage signal (PV093624.1), which is what an actual reagent component looks like, runs the other way. E. coli is the production host of the DNA template plasmids. The paper's conclusion that the DNA is "derived from plasmids used as templates during in vitro transcription of mRNA, rather than from exogenous sources or host genomes" is refuted by the paper's own reads.

The brand split. Every Comirnaty lot sits above every Spikevax lot across a 3,100x range. Source: re-analysis of PRJNA1303864.

2. Whole circular plasmid, not only small fragments. Cross-junction read pairs are nonzero in every lot. The circular-molecule fraction in Comirnaty is two to three orders of magnitude above the original-vector Spikevax lots. The "small fragments" the paper reports describe its own 95 C x 10 min library prep, not the vials.

3. The hybrid mechanism, measured in the deposit's own fragment lengths. This is the finding that closes a loop in this article. Within each lot, same library, same size selection: median spike-region fragment length minus median kanR-region fragment length. Every lot with continuous backbone coverage is positive, +6 to +26 nt; all nine Comirnaty runs sit at +6 to +20 (sign test p = 0.004). The three lots whose coverage collapses at the cassette boundary are the three negative ones (-17, -17, -21). Longer spike-region DNA exactly where spike is enriched is what RNA:DNA hybrid protection predicts (Sutton et al. 1997, quoted above; Lenk et al. 2024, quoted above). The mechanism McKernan proposed and BioNTech's own scientists documented is physically present in the sequencing deposit of the paper published to dismiss it.

The sign flip. Lots with continuous backbone coverage sit at +6 to +26 nt; the three lots whose coverage collapses at the cassette boundary sit at -17 to -21. Source: re-analysis of PRJNA1303864.

4. The kitome objection closed. A minor-allele mixture scan returns zero mixed rows in all five Comirnaty lots and both original-vector Spikevax lots, bounding cross-lot contamination in the sequencing pipeline below 3 percent. The E. coli brand split is not index hopping or library carryover.

One record belongs in the same breath as the findings, because a finding and its funding are read together. The study was commissioned and paid for: approximately EUR 350,000, funded through the Slovak Ministry of Health. The commissioning institute's Biomedical Research Center maintains a documented joint workplace with Sensible Biotech, a commercial mRNA company, and the communicating author, who declared no conflict of interest, is, per that documentation, the institute director whose signature is on the collaboration agreement. The record was assembled by McKernan from public receipts (the academy's annual reports and the Slovak public procurement register) and re-reported by Epoch Times Slovakia in August 2026 from the same primary sources.

What these numbers are not. They are not converted to ng/dose. The only anchors available are the paper's fluorometry and qPCR, and both are the contested methods. Dividing by the reported 8 ng produces a reassuring number by construction, because the divisor is the figure under dispute. The anchors in play, 8 ng reported, 12.2 ng fragmentation-corrected, around 25 ng on the Lander-Waterman treatment, span a factor of three, and any derived dose figure inherits that dispute whole. Sequencing read ratios are, per McKernan's T4-ligase rule, sequencing ratios, not in-vial mass ratios.

Kevin McKernan#

Source: McKernan K. Sequencing of bivalent Moderna and Pfizer mRNA vaccines reveals nanogram to microgram quantities of expression vector dsDNA per dose. OSF Preprint, 2023. doi:10.31219/osf.io/b9t7m

Findings:

• DNA Quantification: qPCR estimates 10-60 ng/dose; fluorometry detected up to 843 ng/dose in some batches
• SV40 Confirmation: Shotgun sequencing identified complete SV40 promoter/enhancer and origin of replication
• Risk Assessment: Highlighted potential for episomal persistence and genomic integration
• Impact: Triggered global replication attempts and regulatory inquiries

Phillip Buckhaults: legislative testimony#

Source: Buckhaults P. South Carolina Senate Medical Affairs Committee Testimony on Pfizer mRNA DNA contamination. Sept 12 2023. Official transcript

What he told the committee:

• Findings: 1-6 ng/dose with SV40 promoter/enhancer confirmation
• Risk Warning: "There is a very real hazard" of the oncogenic promoter in gene therapy contexts
• Integration Status: Not demonstrated in patient samples; exploratory tests only
• Call to Action: Urged manufacturing audits and GMP replication

Additional Laboratory Confirmations#

Brigitte König & Jürgen O. Kirchner (Germany)#

Background: Independent laboratory analysts When: September 2023 What they found: Their analysis revealed the highest contamination levels found to date - 3,600-5,340 ng/dose, representing 360-534 times the regulatory limit. They also confirmed the presence of SV40 promoter sequences. Why it matters: The highest reported levels anywhere, and a serious manufacturing quality-control failure. Research findings: Published in Methods and Protocols

Speicher et al. (Canada)#

Background: Researcher at University of Guelph When: October 2023 (Ontario), March 2024 (Australia), January 2025 (peer-reviewed publication) What he found: Analyzed over 27 vials from different regions and found DNA contamination ranging from 0.22 to 510 ng/dose. His work reported preliminary evidence of a dose-response relationship with serious adverse events, differing between the Pfizer and Moderna products. Peer-reviewed publication: Autoimmunity journal, January 2025. DOI: 10.1080/08916934.2025.2551517, PMID: 40913499. The peer-reviewed version quantified residual plasmid DNA and SV40 sequences exceeding EMA/FDA limits by 36-627-fold across 27 lots, with 815-fold variance between batches indicating severe quality control failures. Why it matters: The analysis spans multiple lots and regions, which is what makes the batch-to-batch variability measurable. Comprehensive reports: Ontario analysis | Australian investigation (archived)

Speicher et al.'s dose-response analysis, which the authors themselves describe as preliminary evidence of a relationship between DNA per dose and the frequency of serious adverse events, reported separately for the Pfizer and Moderna products. Source: OSF preprint mjc97_v1; peer-reviewed version in Autoimmunity 2025;58(1):2551517.

Didier Raoult (France)#

Background: Microbiologist and former professor at Aix-Marseille University When: November 2024 What he found: Using fluorometry and sequencing, Raoult detected 2,712-5,160 ng/dose of complete plasmid DNA including full SV40 promoter/enhancer sequences. Why it matters: Raoult is one of the most cited microbiologists in the world, and his involvement brought the findings to a wider European audience. Study publication: Available through HAL Science Archive

Ulrike Kämmerer (Germany)#

Background: Biologist at University Hospital of Würzburg When: December 2024 What she found: Documented 32.7-43.4 ng/dose of DNA contamination within lipid nanoparticles and confirmed SV40 promoter sequences. She also demonstrated that the contaminated material transfects cells and produces spike protein. Why it matters: The contaminated DNA is biologically active in human cells, not just present. Research publication: Available on ResearchGate - Archived

Source: Kevin McKernan, Laboratory analysis (2024) - Agarose gel electrophoresis showing DNA fragments (~5.3 kb) in mRNA vaccine samples. Lane 1: DNA ladder, Lane 2-4: Various vaccine lots, Lane 5: Negative control

What started as routine quality-control questions turned into an independent investigation across several continents, each laboratory finding the same contamination pattern.

Health Canada's internal emails#

Internal emails obtained through investigative work by Scoops McGoo reveal a regulatory crisis concealed from public view:

The Unplanned Discovery#

• Pfizer's Admission: "Yes, because Pfizer did not identify the presence of SV40 promoter enhancer on the plasmid template used to produce mRNA" (Email Pg. 235)
• Process 2 Connection: This admission directly implicates the commercial manufacturing process where quality controls failed

Regulatory Alarm Bells#

• "SV40 must be avoided!" (Email Pg. 66) - Clear safety principle violation
• "Remedy the situation before Fall 2024 vax campaign" (Pg. 141) - Urgent timeline for Process 2 resolution
• "They do not seem to care much at this moment" (Pg. 151) - Frustration with Pfizer's response

Scientific concern#

• "Fragment size is related to the probability of integration" (Pg. 205), i.e. integration with the human genome
• "References: None" (Pg. 203), sitting beside a "minimal safety risk" assurance

The EMA's Written Concessions and the Public Plasmid Sequence (added 4 October 2026)#

Health Canada's emails are one regulator's internal crisis. The European Medicines Agency's record is now in writing, in its own released document. Under access-to-documents request ASK-289849 (batch 2, release letter EMA/142071/2026, 19 June 2026), the EMA released the BWP position on residual DNA in Comirnaty (EMA/CHMP/BWP/61303/2023, adopted 31 October 2023). Three concessions, verbatim and page-cited. The full treatment, with the EMA document as anchor, is in the companion article The BNT162b2 Plasmid Map.

1. The SV40 elements are in the plasmid. Page 6: "BioNTech/Pfizer has confirmed that SV40 promoter, SV40 Enhancer, SV40 Origin, and SV40 polyA signal elements are present in the DNA plasmid, which is a starting material in the manufacturing process of BNT162b2 / Comirnaty Drug Substance."

2. The manufacturer did not highlight the SV40 content at original authorisation. Page 6: "While the full DNA sequence of the plasmid starting material was provided in the initial marketing authorisation application for Comirnaty, the applicant did not specifically highlight the SV40 sequence, as it was considered to be a non-functional part of the plasmid."

3. The compliance number is not measured on what is in the syringe. Page 5: "Analysis of finished drug product (DP) samples requires additional sample manipulation, including extraction of residual DNA from the LNP, which may impact the ability to sensitively detect and characterize residual DNA." The 10 ng compliance figure is measured on Drug Substance, the upstream intermediate before LNP encapsulation. The Drug Product that is injected is not the material tested.

What the public sequence adds#

The plasmid sequence itself is public: GenBank OR134577.1, McKernan et al.'s deposit, recovered from bivalent vaccine vials in June 2023. The deposit is the pioneers'; the annotation below is this author's, and every coordinate is reproducible from the public record with open tools. The short version of the full eleven-element map:

• 695 bp of SV40 sequence at 100 percent identity to strain 776, in two blocks: a non-coding splice/poly(A) module (positions 165-501, including the 3' terminal 138 nucleotides of the large T-antigen gene, which does not translate) and the full enhancer/promoter/origin block (1096-1453).
• The SV40 promoter is KanR's promoter. The EMA frame, "a non-functional part of the plasmid", is architecturally wrong. The 358 bp SV40 block ends 34 bp upstream of the kanamycin/neomycin resistance CDS and is the designed mammalian promoter for it, inherited unchanged from pCMV-Script (Stratagene, 1997). It is silent in E. coli and functional in any mammalian cell that receives this DNA.
• The vector genealogy is documented to 1981. pSV2-gpt (Mulligan and Berg, PNAS 1981, PMID 7017722) -> pCMV-Script (1997) -> pST1 (Holtkamp et al., Blood 2006, PMID 16940422) -> the pST family -> the BNT162b2 production plasmid. The founding pSV2 paper described transformants that "contain one to five copies of the transfecting DNA associated with, and most probably integrated into, cellular DNA sequences" (pages 2075-2076). Integration was the basis of the transformation system, documented in the primary literature for 45 years. The SV40 content is not a pandemic-era insertion.
• Moderna's plasmid contains zero SV40 sequence (OR134578.1). Same modality, same regulatory pathway, same LNP delivery, a different backbone. SV40 inclusion was a manufacturer design choice, not a technological necessity.
• The BWP's own commitments are unmet in the public record. The position paper committed to two actions by 1 December 2023: analytical evaluation of residual DNA size distribution, intact circular plasmid and SV40 sequences at Drug Substance level, and testing of the residual DNA plasmid's replication capability in bacteria. Confirmatory finished-product testing by the manufacturer or an OMCL "could also be considered", which is the regulator's way of saying it is not required.

Source: this author's annotation of GenBank OR134577.1 (McKernan et al. deposit), companion paper DOI 10.5281/zenodo.21918632. Full element table in The BNT162b2 Plasmid Map.

The category error: mRNA safety is not residual DNA safety#

A pattern runs through the regulatory and publishing response to this contamination record. Safety arguments constructed for the mRNA product are applied to the residual plasmid DNA as if the two molecular species were interchangeable. They are not. The mRNA is the intended product: cytoplasmic, non-integrating, degraded. The residual plasmid DNA is a process contaminant and a different molecular species: double-stranded, carrying the SV40 enhancer/promoter, the KanR/NeoR cassette and the bacterial origin, stable in ways the mRNA is not, and transportable into the nucleus. A safety case that begins and ends with the properties of the mRNA does not address the properties of the contaminant. Polykretis et al. (Oncotarget, 2026) document the error with two verbatim peer-review rejection letters, both of which argue from the mRNA's non-integration to the safety of the residual DNA.

What we still do not know#

The contamination is confirmed. What it means biologically is not, and the questions below are barely studied.

Complete research agenda: Critical Research Gaps: What We Still Don't Know About DNA Contamination covers:

• Nucleic acid sensing & immune evasion (TLR7/8, cGAS-STING, N1-methylpseudouridine)
• Persistence & biodistribution (DNA half-life in tissues, LNP effects)
• Genomic integration risk (LINE-1 mechanisms, cell cycle effects)
• Dose, timing & threshold effects (cumulative risk, dose-response)
• Individual variability & susceptibility (genetic polymorphisms, inflammation)
• Concomitant factors (medications, prior infections, nutritional status)
• Manufacturing & quality control (batch variance, SV40 consistency)
• Long-term surveillance & biomarkers (delayed effects, registries)

The five sharpest open items: whether DNA levels correlate with adverse events, why some lots show minimal contamination while others exceed limits 500-fold, where plasmid DNA travels and for how long, whether population-level effects exist, and what data supports the "minimal risk" determination.

Health Canada's internal emails said "SV40 must be avoided" and sought to "remedy the situation before Fall 2024" while the public position stayed reassuring (see the Health Canada section). These questions were answerable before billions of doses were given. They were instead called "theoretical concerns".

Process 1 vs Process 2: the unannounced switch#

The clinical-trial material and the commercial material were made differently.

• Process 1: the clinical trial material, the version tested for safety and efficacy in the human trials
• Process 2: the commercial scale-up, with different purification, different enzymes, and different DNA template production

The switch to Process 2 for global supply was not publicly announced. It mattered because Process 2 used different materials and methods, the ones later found to produce the higher contamination levels. Pfizer's original protocol included a direct Process 1 vs Process 2 comparison cohort, the standard requirement when a manufacturing process changes. Protocol Amendment 20 (September 2022) removed that comparison as Process 2 dominated supply (Archive of Protocol C4591001 Amendment 20). No formal comparison was ever published, so there is no public data demonstrating manufacturing equivalence.

When researchers and citizens asked regulatory agencies for evidence that the two processes were equivalent, the UK MHRA acknowledged the question in its response to FOI 23/510 but released no analytical dossier.

For detailed analysis of manufacturing changes and their implications, see Pfizer Process 1 vs Process 2: What Changed, What's Measured, and What Remains Uncertain

Figure 6: mRNA vaccine manufacturing process flowchart comparing Process 1 (enzymatic/PCR-based template preparation) and Process 2 (E. coli plasmid DNA template amplification). The critical divergence point occurs at the DNA template preparation stage. Process 2 introduces bacterial DNA contamination including SV40 promoter sequences that can contaminate the final mRNA product through incomplete DNase treatment and RNA:DNA hybrid formation. Source: Process and analytical strategies for the safe production of mRNA vaccines and therapeutics, PMC12819531

Somatic hypermutation: the SV40 enhancer finding#

Šenigl and colleagues found that the SV40 enhancer functions as a somatic hypermutation targeting element. The finding appeared as a bioRxiv preprint in January 2024 (DOI: 10.1101/2024.01.09.574829, PMID: 38260396) and is now peer-reviewed: Šenigl et al., Tumour Virus Res 2024;18:200293, PMID 39490533, "The SV40 virus enhancer functions as a somatic hypermutation-targeting element with potential tumorigenic activity."

What is Somatic Hypermutation?#

Somatic hypermutation (SHM) is a natural biological process that's essential for healthy immune function:

1. B-cells are white blood cells that produce antibodies to fight infections
2. When B-cells encounter new pathogens, they undergo SHM to fine-tune their antibodies
3. This process creates antibody diversity, allowing our immune system to recognize millions of different pathogens
4. The AID enzyme (Activation-Induced Cytidine Deaminase) acts as the "master catalyst" for this process

The SV40 Concern#

The research suggests that SV40 enhancer sequences might redirect or dysregulate this natural mutation process. If SHM does not stay on its normal targets:

Potential Consequences:

1. Impaired Antibody Production: If SHM doesn't work correctly, the immune system might not be able to produce effective antibodies against new infections
2. Unexpected DNA Mutations: If the mutation process goes awry, it could potentially affect other parts of the genome
3. Autoimmune Issues: Misdirected hypermutation could theoretically contribute to autoimmune conditions

Why this matters scientifically#

As McKernan and Speicher's analyses show, compromised antibody responses indicate disrupted immune function.

The concern isn't that SV40 definitely causes these problems, but that introducing a powerful genetic enhancer sequence into the body creates theoretical risks that deserve thorough investigation.

Cryptic promoters: expression without integration#

A second promoter risk sits in the bacterial origin of replication itself: it creates biological activity with no genomic integration required.

Lemp et al. (2012)#

Lemp et al. (Nucleic Acids Research, 2012, PMID 22618870) demonstrated that ColE1/pUC origins, the same origins used in the vaccine plasmids, contain cryptic mammalian promoters that drive read-through transcription.

Key Findings:

• ColE1/pUC bacterial origins contain cryptic promoter sequences
• These promoters are active in mammalian cells
• They drive transcription of downstream sequences
• No genomic integration required: episomal expression is sufficient

DOI: 10.1093/nar/gks451

Human Evidence: Ryan et al. RNA-Seq#

Ryan et al. performed RNA-Seq analysis on blood from vaccinated individuals and found elevated coverage over the ori region (bp 2890-3478), the location where Lemp identified cryptic promoter activity.

This is in vivo transcription evidence from vaccinated humans.

Why This Changes the Risk Model#

The traditional integration risk model: the plasmid must integrate into the genome to drive expression, and integration is rare, so the risk is low.

The cryptic promoter model: episomal DNA in the nucleus drives expression without integration, so the risk does not depend on a rare event. LNPs deliver DNA to the nucleus, and episomal persistence is documented.

The dual pathway:

Two independent promoter systems are documented in vaccine plasmids:

Both can drive unintended gene expression from residual plasmid DNA. SV40 actively promotes nuclear entry and integration. ColE1 cryptic promoter works from episomal DNA that persists in the nucleus.

Regulatory Blind Spot#

Current regulatory assessments assume that without integration, residual DNA is inert. The cryptic promoter data contradict the assumption:

• Assumption: No integration = no biological activity
• Reality: Cryptic promoters + episomal persistence = ongoing expression
• Gap: Regulatory safety assessments never considered integration-independent expression

This creates a pathway for residual DNA to be biologically active that manufacturers and regulators never addressed in risk assessments.

Biological Consequences#

If residual plasmid DNA with cryptic promoters persists in cell nuclei:

1. Persistent antigen expression: Ongoing Spike production without integration
2. Off-target transcription: Unintended sequences expressed
3. Chronic immune stimulation: Foreign DNA continuously detected
4. Dose uncertainty: No way to quantify duration or amount of expression

None of these scenarios were studied in clinical trials or addressed in regulatory submissions.

Known before the pandemic#

Lemp's paper predates it: the cryptic promoter activity of these bacterial origins was in the molecular biology literature for years before the vaccine plasmids were designed. Manufacturers and regulators had access to that literature. It was never addressed in mRNA vaccine risk assessments or discussed in regulatory review documents.

References:

• Lemp et al. "Cryptic promoter activity in the ColE1/pUC origin of replication." Nucleic Acids Research, 2012. PMID: 22618870, DOI: 10.1093/nar/gks451
• Ryan et al. RNA-Seq analysis of vaccinated blood samples (cited in McKernan Substack, December 2024)

Lemp et al. 2012, PMID 22618870: https://pubmed.ncbi.nlm.nih.gov/22618870/

Manufacturer prior knowledge: the Moderna patents#

Moderna's own intellectual property shows the company understood the residual-DNA problem in mRNA manufacture years before COVID-19. The primary document is US 10,077,439 B2, "Removal of DNA fragments in mRNA production process": ModernaTx, Inc., priority date March 15, 2013, filed March 13, 2014. The patent is itself a method for removing DNA fragments from the mRNA production process; it exists because DNA removal is a recognized requirement of the process.

What the patent says#

Verbatim, from the specification:

• On why the DNA template must be removed: "residual DNA in drug products may induce activation of the innate response" and "has the potential to be oncogenic in patient populations." The same passage states "Regulatory guidelines may also require the quantification, control, and removal of the DNA template in RNA products."
• On qPCR's blind spot: "Quantitative PCR is often applied to measure the residual DNA," but "it only detects the DNA molecules that contain both qPCR primers" and does not measure "all other smaller DNA molecules that are partially digested." In Moderna's own filing, the standard method was acknowledged to undercount fragmented DNA. That is the same measurement gap the independent laboratories described above ran into when regulators relied on qPCR to clear these lots.
• On the production fix: "1 U of RNase-free DNase is then used to digest the original template," with the digested DNA then separated from the RNA transcript.

A separate Moderna filing, US 2019/0240317 A1 (HPIV3 RNA vaccines, priority date 2015, filed March 2019), discusses genomic integration in its background section, warning that DNA vaccination carries "the possibility of insertional mutagenesis," "which could lead to the activation of oncogenes or the inhibition of tumor suppressor genes," and positioning RNA vaccines as avoiding that risk. Read the context precisely: that passage describes injected plasmid DNA vaccines, not residual DNA contamination in mRNA manufacture. It documents that Moderna's patent drafters understood insertional mutagenesis as a DNA-vaccine risk category; it is not a contamination admission.

Correction, August 2026: an earlier version of this section presented four statements ("Because the remaining DNA fragments may cause a patient to develop cancer"; "Some of the introduced DNA may be incorporated into the genome of the cell and inherited by the offspring"; and two others) as direct patent quotations. They were paraphrases circulating in social-media graphics, not patent text; US 10,077,439 B2 contains neither the word "cancer" nor "offspring." The patent's actual language is now quoted above, character-for-character, from the primary document. The substance of the manufacturer-knowledge argument stands on the verified text alone.

Why this knowledge matters#

1. Moderna documented residual DNA as a safety concern in its own patent literature, years before the pandemic
2. It stated oncogenic potential in plain language
3. It taught that qPCR undercounts partially digested DNA, the same method later used to clear commercial lots
4. Its patent drafters understood insertional mutagenesis as a live risk category for injected DNA

The timing#

The US 10,077,439 B2 family predates the pandemic (priority 2013), and the integration language in the 2019 application was filed in March 2019, before SARS-CoV-2 emerged. So manufacturers had documented knowledge of residual-DNA risks in mRNA technology in their own filings, had taught in patent text that qPCR does not measure partially digested DNA fragments, were aware of oncogenic and integration concerns as risk categories, and did not volunteer any of this during the vaccine rollout.

Biological risk assessment#

Risk mechanisms#

1. Genomic integration: foreign DNA incorporation with potential oncogene activation
2. Promoter-driven expression: SV40 enhancer activating nearby genes
3. Inflammatory response: foreign DNA triggering immune reactions via cGAS-STING
4. Somatic hypermutation disruption: SV40 targeting of the AID enzyme and antibody diversity machinery

Evidence status#

• In vitro confirmation: multiple studies show SV40 promoter activity and DNA presence
• In vivo gap: no demonstrated integration or clinical sequelae in vaccinated populations
• Biological plausibility: multiple identified mechanisms with historical precedent

Coverage where the mechanisms point. Nanopore reads recovered from vaccine vials map across the spike ORF and the SV40 promoter/enhancer of the production plasmid: the coding sequence the integration and expression mechanisms above turn on, and the enhancer/promoter that can drive transcription in any mammalian cell that takes up the DNA. Read-level detection is exposure evidence, not integration evidence. Source: McKernan et al., "RNA:DNA hybrids survive digestion in mRNA vaccine manufacturing", Zenodo 2025, doi:10.5281/zenodo.17832183.

Can mRNA Integrate Into Human DNA?#

For decades, the conventional understanding was that mRNA vaccines couldn't integrate into human DNA because mRNA is temporary and doesn't enter the cell nucleus. However, recent research has challenged this assumption.

Key Study: In 2022, Aldén et al. published findings that BNT162b2 mRNA could be reverse-transcribed into DNA in human liver cells (Current Issues in Molecular Biology).

What This Means:

• Reverse transcription is the process where RNA is converted back into DNA
• This can occur through natural enzymes like LINE-1 retrotransposons that exist in human cells
• If mRNA can be converted to DNA, it raises questions about whether DNA contamination could integrate into the genome
• This challenges the long-standing assumption that mRNA vaccines cannot affect human DNA

Immune System Changes#

Researchers have documented several unexpected immunological effects following mRNA vaccination:

IgG4 Class Switching:

• A 2023 study by Irrgang et al. found that repeated mRNA vaccination causes the immune system to shift toward producing IgG4 antibodies (PubMed)
• IgG4 antibodies are typically less inflammatory and less effective at fighting infections
• This class switching is unusual and its long-term implications are not fully understood

Spike Protein Persistence:

• Patterson et al. detected SARS-CoV-2 spike protein in immune cells up to 15 months after infection (Frontiers in Immunology)
• While this study focused on natural infection, it raises questions about how long vaccine-produced spike protein might persist

Medical Findings After Vaccination#

Several published autopsy studies have reported unusual findings after mRNA vaccination:

Brain Inflammation:

• Mörz (2022) documented multifocal necrotizing encephalitis (brain inflammation) following BNT162b2 vaccination (Vaccines)

Heart Muscle Damage:

• Schwab et al. characterized myocarditis pathology following vaccination (PMC)

Young Adult Deaths:

• Gill et al. reported autopsy findings in two teenagers who died after Pfizer vaccination (Archives of Pathology & Laboratory Medicine - Archived)

Important Scientific Context#

Four qualifications:

1. Correlation vs. Causation: These findings don't prove that DNA contamination caused these outcomes
2. Individual Cases: Autopsy reports typically involve individual cases, which may not represent population-level effects
3. Alternative Explanations: Other factors could contribute to these observed effects
4. Research Ongoing: The scientific community is still investigating these mechanisms

The open question#

DNA contamination carrying an SV40 enhancer raises three questions:

• Could it contribute to unexpected biological effects?
• Were these possibilities investigated before widespread use?
• Does biologically active genetic material change the safety profile?

All three deserve thorough, independent investigation.

Regulatory Standards vs Reality#

Established Safety Benchmarks#

• WHO TRS 978 Annex 3: ≤10 ng residual DNA per dose with <200 bp fragmentation
• ICH Q5D/Q6B: Validated purification and risk assessment requirements
• Historical Context: Standards developed from gene therapy and viral vector experience

The Transparency Gap#

• Regulator Statements: Health Canada, EMA, and FDA maintain authorized lots meet specifications
• Missing Data: No public release of raw lot testing data or comprehensive comparability studies
• Independent Verification: No GMP-grade replication despite multiple lab confirmations

What should happen next#

Immediate Needs#

1. Full Data Transparency: Release all manufacturing quality control and comparability data
2. Multi-Laboratory Audit: GMP-grade independent testing of retained lots
3. Clinical Correlation: Examine adverse events by manufacturing lot and DNA levels

Long-Term Solutions#

4. Regulatory Reform: Emergency use products should match conventional medicine quality standards
5. Independent Oversight: Separate safety monitoring from promotional review
6. Manufacturing Standards: Updated guidelines for mRNA product purity and characterization

Related investigations#

• The BNT162b2 Plasmid Map: What EMA Confirmed and What the Public Sequence Shows The full eleven-element annotation of OR134577.1, the 695 bp SV40 footprint, the 1981 vector genealogy, the redacted dossier section, and the EMA BWP position paper quotes in full. Companion paper on Zenodo.

• The Host Genome They Said Was Not There: The Achs Sequencing Deposit Against the Achs Paper The complete read-level re-analysis of PRJNA1303864: the run table for all fifteen libraries, the E. coli genome screen, the plasmid topology work, the fragment-length contrast, the mixture scan, and the standing caveats.

• The Stability Trap: How mRNA Vaccine Engineering and DNA Contamination Created a Perfect Storm Analysis of how mRNA stability modifications intersect with DNA contamination to create biological risks.

• Pfizer Process 1 vs Process 2: What Changed, What's Measured, and What Remains Uncertain Detailed investigation into manufacturing changes between clinical trial and commercial vaccine lots.

What This Article Does Not Claim (added 4 October 2026)#

• No SV40 large T-antigen protein. The sequence evidence excludes T-ag translation from OR134577.1: no start codon upstream of the T-ag fragment, no promoter that would transcribe it, no splice donor and acceptor pair that could assemble a T-ag mRNA.
• No whole SV40 virus. Under 7 percent of the T-antigen gene, no capsid DNA, no SV40 miR-S1.
• No deliberate pandemic-era SV40 insertion. The SV40 content is a 1981 legacy of the pSV2 vector family, inherited through pCMV-Script (1997) into the pST family (2006).
• No demonstrated genomic integration in vaccinated people. The integration property of the vector family is documented in its founding literature; observed integration in patients is not established.
• No ng/dose conversion from sequencing read shares. The measurement anchors are contested (see the Achs section); any conversion inherits the dispute.
• Detection is not persistence of function, and persistence is not causation. The contamination is established. The clinical sequelae are not.

Conclusion#

The SV40 DNA contamination record is a cascade of failures:

• Manufacturing control: unexpected SV40 sequences in final products
• Regulatory transparency: internal alarm without public disclosure
• Scientific integrity: removal of planned safety comparisons
• Corporate accountability: prior knowledge of risks without adequate mitigation

The burden of proof now rests with manufacturers and regulators to demonstrate that these findings truly represent "minimal risk" to public health. These findings are confirmed across eight independent laboratories, corroborated by the regulators' own testing, and supported by identified biological mechanisms.

Update, 4 October 2026. The record has moved since this article was written. The regulator's own released documents now concede in writing what the laboratories found: the EMA BWP position paper confirms the four SV40 elements in the plasmid starting material and the fact they were not highlighted at authorisation, and measures compliance on Drug Substance rather than the injected product. The paper commissioned to rebut the contamination findings has been contradicted by its own deposited sequencing data. What the record still does not contain: mandated finished-product testing by any regulator, a GMP-grade independent audit of retained lots, or demonstrated genomic integration in vaccinated people. The contamination is established. The clinical questions were the questions that mattered all along, and they remain open.

Key References#

RNA:DNA Hybrid Mechanism:

• Sutton DH, Conn GL, Brown T, Lane AN. "The dependence of DNase I activity on the conformation of oligodeoxynucleotides." Biochem J. 1997;321(Pt 2):481-486. PMID: 9020884. (DNA strands of RNA:DNA hybrid duplexes are not cleaved by DNase I)

DNA Contamination Studies:

• McKernan K, et al. "RNA:DNA hybrids survive digestion in mRNA vaccine manufacturing." Zenodo. 2025. doi:10.5281/zenodo.17832183
• Speicher DJ, Rose J, McKernan K. "Quantification of residual plasmid DNA and SV40 promoter-enhancer sequences in Pfizer/BioNTech and Moderna modRNA COVID-19 vaccines from Ontario, Canada." Autoimmunity. 2025;58(1):2551517. DOI: 10.1080/08916934.2025.2551517, PMID: 40913499

Regulatory Documents:

• EMA Assessment Report REC 027 (2021). Required DNase stability studies.
• Health Canada FOI emails (2023). SV40 contamination internal communications.

Regulatory and Sequence Records (added 4 October 2026):

• EMA/CHMP/BWP/61303/2023. BWP position on residual DNA in Comirnaty. Adopted 31 October 2023. Released under ASK-289849 batch 2 (release letter EMA/142071/2026, 19 June 2026).
• GenBank OR134577.1 (BNT162b2 production plasmid, McKernan et al. deposit, June 2023); GenBank OR134578.1 (mRNA-1273 production plasmid, SV40-free).
• Mulligan RC, Berg P. "Selection for animal cells that express the Escherichia coli gene coding for xanthine-guanine phosphoribosyltransferase." Proc Natl Acad Sci USA. 1981;78(4):2072-2076. PMID 7017722.
• Holtkamp S, Kreiter S, Selmi A, et al. "Modification of antigen-encoding RNA increases stability, translational efficacy, and T-cell stimulatory capacity of dendritic cells." Blood. 2006;108(13):4009-4017. PMID 16940422.
• Emmett T. Independent sequence-derived annotation of the BNT162b2 plasmid starting material: vector genealogy, SV40 footprint, and the 1981 integration property. Zenodo. DOI 10.5281/zenodo.21918632 (concept).

Achs et al. and Its Critique (added 4 October 2026):

• Achs et al. "Systematic analysis of COVID-19 mRNA vaccines using four orthogonal approaches demonstrates no excessive DNA impurities." npj Vaccines. 2025;10:259. DOI 10.1038/s41541-025-01304-9. Sequencing deposit: PRJNA1303864.
• McKernan K, Speicher D, Rose J. "Systematic methodological flaws in DNA contamination assessment of mRNA vaccines: a critical analysis of Achs et al. (2025)." Oncotarget. 2026;17:368-376.
• Šenigl F, Soikkeli AI, Prost S, Schatz DG, Slavková M, Hejnar J, Alinikula J. "The SV40 virus enhancer functions as a somatic hypermutation-targeting element with potential tumorigenic activity." Tumour Virus Res. 2024;18:200293. PMID 39490533. PMC11564006. (Peer-reviewed version of the SHM finding cited in the mechanism section.)
• Polykretis P, Seneff S, Tran K, et al. "The censorship of post-COVID-19 injection residual DNA biosafety research: a case report." Oncotarget. 2026. (The two rejection letters quoted in the category-error discussion.)

This investigation continues because independent scientists, journalists, and concerned citizens refuse to accept "trust us" as adequate scientific evidence for products administered to billions of people.